Hedging weather risk in agriculture
Abstract
Hedging revenues has been practiced for decades. Through the use of futures, forwards, swaps and options, companies have been able to hedge their associated risks. Historically, the process of hedging has mainly been related to price, while volumetric risks have not been considered to an equal extent. As the prices in Norwegian agriculture are set at a national level annually, the main source of risk relates to volume. The purpose of this thesis is to analyse the effect of hedging volumetric risk on the income of Norwegian farmers, by using weather derivatives. The effects are measured in both change of volatility and change in a simplified Sharpe ratio.
Weather derivatives are relatively underexplored compared to more traditional financial instruments and provide a new way of managing risks in industries that are strongly dependent on weather patterns. Agriculture relies heavily on weather conditions, and temperature, precipitation and sunlight all directly influence crop yields. This dependency makes the sector vulnerable to weather-related risks. Weather derivatives provide solutions to hedging against volumetric risks, such as fluctuating crop yields due to unpredictable weather, offering farmers a way to mitigate some of these uncertainties. After consulting with an expert in the field, we received clear indications that, in Norway, the primary weather risk is related to precipitation. While temperature and sunlight hours undeniably affect crops, their primary effect in the Norwegian climate is limited to influencing the time required for the grain to be ready for harvest. With this knowledge in mind, the thesis aims to construct precipitation derivatives that can help farmers stabilize their income. Agriculture is generally characterized by thin margins, and large fluctuations may result in substantial financial concerns.
As weather derivatives do not have a tradeable asset as the underlying, traditional methods to price financial derivatives, such as the Black-Scholes model, are not applicable. The majority of weather derivatives are traded over-the-counter (OTC) and consequently there is little data available, especially in Norway, on how to price them. This thesis will use some of the most common pricing methods that occur in previous literature – the historical burn analysis and Monte Carlo simulation. Through regression analysis, we aim to quantify the effect of precipitation on crop yields and use this as a basis for designing the derivative contracts. The findings of the analysis show that applying the proposed hedging strategy on an out-of-sample dataset through backtesting yielded an improvement in the majority of organization numbers.